서브메뉴
검색
Domain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data
Domain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105232
- ISBN
- 9798297610064
- DDC
- 621.3
- 저자명
- McCrabb, Andrew.
- 서명/저자
- Domain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data
- 발행사항
- [Sl] : University of Michigan, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 214 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-04, Section: B.
- 주기사항
- Advisor: Bertacco, Valeria M.
- 학위논문주기
- Thesis (Ph.D.)--University of Michigan, 2025.
- 초록/해제
- 요약Graph-based processing enables many applications in logistics, e-commerce, social media, and more. However, graph workloads are slow: they are bottlenecked not by compute power, but by inefficient data access. As useful graphs get larger and graph-based algorithms become more complex, adding more powerful compute units like Graphics Processing Units (GPUs) or Tensor Processing Units (TPUs) cannot keep up with the increasing size and complexity of these workloads.To address these challenges, we first introduce a novel taxonomy of graph-based algorithms: those that treat graphs (1) as data frameworks, (2) as algorithmic frameworks, or (3) as both. Each category has overlapping needs: higher memory bandwidth, better data organization, and greater thread-level parallelism. Next, we demonstrate that custom processing-in-memory (PIM) hardware accelerators are effective and energy-efficient solutions to the compute and memory bottlenecks of graph-based applications.Specifically, we propose and evaluate three custom PIM accelerators, DREDGE (for graph-as-data-framework applications), ACRE (for graph-as-algorithmic-framework applications), and GLEAM (for graph-as-both applications), each targeting one of the three categories of graph applications. DREDGE targets dynamic graph applications by introducing a novel partitioning technique and dedicated hardware support to continuously improve data organization in memory. ACRE accelerates the training of tree-based machine learning models in a way that allows users to better understand the models' reasoning. GLEAM targets graph neural networks, the primary machine learning models for graph-based data, accelerating the node aggregation operations that bottleneck training and inference operations. These three designs offer a 2.5-14x speedup for their respective applications, and they save 77-93% of total system energy over their respective baselines. Each design fits within the logic area of modern 3D-stacked memory: 0.3-13% of the available logic space. Finally, we present two benchmark suites, DyGraph and BeXAI, making them publicly available to support future research in dynamic graphs processing and explainable machine learning acceleration. Together, these contributions enable efficient and scalable graph computing to handle the demands of tomorrow's graph workloads.
- 일반주제명
- Computer engineering
- 일반주제명
- Computer science
- 일반주제명
- Information technology
- 키워드
- Memory
- 키워드
- Graph processing
- 기타저자
- University of Michigan Computer Science & Engineering
- 기본자료저록
- Dissertations Abstracts International. 87-04B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017359894
■00520260202105232
■006m o d
■007cr#unu||||||||
■020 ▼a9798297610064
■035 ▼a(MiAaPQ)AAI32271908
■035 ▼a(MiAaPQ)umichrackham006396
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a621.3
■1001 ▼aMcCrabb, Andrew.
■24510▼aDomain-Specific Benchmarks and Architectures for Applications Using Graph-Based Data
■260 ▼a[Sl]▼bUniversity of Michigan▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a214 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-04, Section: B.
■500 ▼aAdvisor: Bertacco, Valeria M.
■5021 ▼aThesis (Ph.D.)--University of Michigan, 2025.
■520 ▼aGraph-based processing enables many applications in logistics, e-commerce, social media, and more. However, graph workloads are slow: they are bottlenecked not by compute power, but by inefficient data access. As useful graphs get larger and graph-based algorithms become more complex, adding more powerful compute units like Graphics Processing Units (GPUs) or Tensor Processing Units (TPUs) cannot keep up with the increasing size and complexity of these workloads.To address these challenges, we first introduce a novel taxonomy of graph-based algorithms: those that treat graphs (1) as data frameworks, (2) as algorithmic frameworks, or (3) as both. Each category has overlapping needs: higher memory bandwidth, better data organization, and greater thread-level parallelism. Next, we demonstrate that custom processing-in-memory (PIM) hardware accelerators are effective and energy-efficient solutions to the compute and memory bottlenecks of graph-based applications.Specifically, we propose and evaluate three custom PIM accelerators, DREDGE (for graph-as-data-framework applications), ACRE (for graph-as-algorithmic-framework applications), and GLEAM (for graph-as-both applications), each targeting one of the three categories of graph applications. DREDGE targets dynamic graph applications by introducing a novel partitioning technique and dedicated hardware support to continuously improve data organization in memory. ACRE accelerates the training of tree-based machine learning models in a way that allows users to better understand the models' reasoning. GLEAM targets graph neural networks, the primary machine learning models for graph-based data, accelerating the node aggregation operations that bottleneck training and inference operations. These three designs offer a 2.5-14x speedup for their respective applications, and they save 77-93% of total system energy over their respective baselines. Each design fits within the logic area of modern 3D-stacked memory: 0.3-13% of the available logic space. Finally, we present two benchmark suites, DyGraph and BeXAI, making them publicly available to support future research in dynamic graphs processing and explainable machine learning acceleration. Together, these contributions enable efficient and scalable graph computing to handle the demands of tomorrow's graph workloads.
■590 ▼aSchool code: 0127.
■650 4▼aComputer engineering
■650 4▼aComputer science
■650 4▼aInformation technology
■653 ▼aMemory
■653 ▼aGraph processing
■653 ▼aHardware accelerators
■653 ▼aGraphics Processing Units
■653 ▼aTensor Processing Units
■690 ▼a0464
■690 ▼a0984
■690 ▼a0489
■71020▼aUniversity of Michigan▼bComputer Science & Engineering.
■7730 ▼tDissertations Abstracts International▼g87-04B.
■790 ▼a0127
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17359894▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


